Convenient and versatile synthesis of formyl-substituted benzoxaboroles

Article abstract of DOI:10.1016/j.tet.2009.08.026

Despite of the medicinal significance of benzoxaboroles, with the newly discovered clinical compound AN2690 as an example, the synthetic method for rapid diversification of this novel scaffold is lacking. To this end, a versatile and scalable synthesis of formyl-substituted benzoxaboroles is described here. A key step is the mono-oxidation of the two hydroxyls in compound 4 by taking advantage of the stable oxaborole ring in non-coordinating solvents, which was devised based on the study of the intramolecular coordination and exchange properties.

Full text of DOI:10.1016/j.tet.2009.08.026

Tetrahedron 65 (2009) 8738–8744
Contents lists available at ScienceDirect
Tetrahedron
journal homepage: www.elsevier.com/locate/tet
Convenient and versatile synthesis of formyl-substituted benzoxaboroles
Long Ye y, Dazhong Ding, Yiqing Feng z, Dongsheng Xie, Puhua Wu, Hui Guo,
*
Qingqing Meng, Huchen Zhou
School of Pharmacy, Shanghai Jiao Tong University, 800 Dong Chuan Road, Shanghai 200240, China
a r t i c l e i n f o
a b s t r a c t
Article history:
Received 2 April 2009
Received in revised form 10 August 2009
Accepted 11 August 2009
Available online 13 August 2009
Despite of the medicinal significance of benzoxaboroles, with the newly discovered clinical compound
AN2690 as an example, the synthetic method for rapid diversification of this novelscaffold is lacking. To this
end, a versatile and scalable synthesis of formyl-substituted benzoxaboroles is described here. A key step is
the mono-oxidation of the two hydroxyls in compound 4 by taking advantage of the stable oxaborole ring in
non-coordinating solvents, which was devised based on the study of the intramolecular coordination and
exchange properties.
Ó 2009 Elsevier Ltd. All rights reserved.
1. Introduction
a tetrahedral covalent adduct with the cis-diol on tRNA, which re-
sults in trapping of tRNA in the enzyme’s editing domain.3 As
Boron, neighbouring carbon in the periodic table, has been
a largely under-utilized element in modern medicinal chemistry
research. Although boronic acids have been investigated as enzyme
inhibitors of proteases, proteasome, arginase, nitric oxide synthase,
esterase and transpeptidase during the past two decades,1 none of
them had become a clinically usable drug until bortezomib (Fig. 1),
a peptidyl boronic acid that inhibits 20S proteasome, was approved
by US FDA in 2003 for the treatment of multiple myeloma,2 which
inspired enthusiasm in boron-containing molecules as a novel
druggable chemical class. As an extension of boron’s new role in
drug discovery, AN2690 (Fig. 1), a small benzoxaborole, was dis-
covered to inhibit leucyl-tRNA synthetase (LeuRS) by taking ad-
vantage of its readily sp2–sp3 inter-convertible boron atom to form
a treatment against onychomycosis, AN2690 is currently under
clinical trials.4 Despite of the surging medicinal significance of
benzoxaboroles, the study of their synthesis and properties has
been very limited4–7 and the synthetic method for rapid di-
versification of this novel structure is highly desired. Thus, we
undertook the task to devise a convenient and versatile synthetic
method for the rapid diversification of the benzoxaborole scaffold
in order to expand its utilization in drug discovery.
In this report, we describe the development of a practically
scalable synthesis of formyl-substituted benzoxaboroles as highly
versatile and medicinally useful scaffolds and their derivatization.
First, 2,6-diformylbromobenzene (3), which is also a useful building
block in supramolecular chemistry, was obtained using a mild and
readily scalable method (Scheme 2). The intramolecular co-
ordination phenomenon of compounds 14 and 15 (Figs. 2 and 3),
and the effect of solvents on exchange rate was investigated. The
empty p-orbital possessed by the boron atom offered a unique
protective group for its neighbouring hydroxyl group, thus made
possible the quantitative mono-oxidation of the two equivalent
hydroxyl groups in non-coordinating solvent. We thus developed
a mild and scalable synthesis of 7-formylbenzoxaborole (1) with
47.3% overall yield over six steps from the commercially available
2,6-dimethylbromobenzene (6). Furthermore, this synthetic route
can be readily applied to any substitution position (4-, 5-, 6-, or 7-)
on the benzoxaborole’s aromatic ring, which for the first time
provided access to the systematic modification of the benzox-
aborole scaffold. Combined with the chemical versatility of the
formyl group, we thus established a valuable synthetic tool for the
study of benzoxaboroles as potential drug candidates. The sub-
sequent derivatization reactions were investigated and proved to
be compatible with the benzoxaborole functionality.
OH
7
4
O
OH
B
H
N
B
1
O
6
N
N
N
H
OH
2
F
O
5
3
Bortezomib (Velcade, PS-341)
AN2690
Figure 1. Structures of bortezomib and AN2690.
* Corresponding author. Tel.: þ86 21 3420 6721; fax: þ86 21 3420 4457.
E-mail address: hczhou@sjtu.edu.cn (H. Zhou).
y
Present address: College of Life Science and Biotechnology, Shanghai Jiao Tong
University, 800 Dong Chuan Road, Shanghai 200240, China.
z
Present address: Department of Chemistry, Pennsylvania State University,
University Park, PA 16802, USA.
0040-4020/$ – see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.tet.2009.08.026

L. Ye et al. / Tetrahedron 65 (2009) 8738–8744
8739
A
in CD3OD
O
O
CHO
d
x
Br
x
Br
c
a
b
d'
a'
c'
b'
CHO
CHO
CHO
2
OH
3
Br
B
O
OH
OH
5.00
4.50
4.00
3.50
3.00
2.50
2.00
1.50
1
6
OH
Br
B
in CD3OD+Na
B
O
c
OH
a
d
b
4
5
x
Scheme 1. Retrosynthetic route A and B towards 1.
2. Results and discussion
5.00
4.50
4.00
3.50
3.00
2.50
2.00
1.50
In order to explore benzoxaboroles as a medicinally useful scaf-
fold, we embarked on the task to device a convenient and scalable
synthesis of formylbenzoxaboroles that can be converted to a variety
of different functionalized benzoxaboroles. We aimed to develop
a one-fit-all method that can be easily applied to any of the four
positions on the aromatic ring, namely 4-, 5-, 6- or 7- position (Fig.1).
At the same time, we were urged to explore a route that is amenable
to rapid scale-up with mild conditions and facile purifications.
Using 7-formylbenzoxaborole as an example, the retrosynthetic
analysis is shown (Scheme 1). Dimethylbromobenzenes have all
four possible dimethyl substitution patterns (2,6-, 2,5-, 2,4- and
2,3-) commercially available and were chosen as the starting ma-
terials. The key was to break down the equivalency of the two
methyl groups, rendering one into hydroxymethyl to close the
oxaborole ring while the other into formyl oxidation state with
satisfactory yields.
HO
C
c
c'
OH
b'
CD3
OH
b
D
a'
O
B
a
-CD3OD
+CD3OD
B
OH
O
O
d'
d
HO
NaOCD3
OCD3
NaOCD3
c
b
OCD3
a
B
O
Na+
d
Figure 2. Top: (A) Compound 14 in CD3OD existed as two components. Ha, Hb and Hc
0
0
0
from free propoxyl; Ha , Hb and Hc from boron-coordinated propoxyl. (B) Addition of
sodium gave a single component. Ha, Hb and Hc from free propoxyl. (C) Proposed
equilibrium of compound 14 in CD3OD and single component after addition of sodium.
COOMe
Br
COOMe
Br
NaOH
We tried to differentiate the two methyl groups in their carboxy-
late or formyl oxidation states with moderate success, and they were
not applicable to large scale synthesis due to low yields and re-
quirement of column chromatography in purification. For example,
treatment of diester 7 with NaOH gave the mono-acid 8 in 79% yield
(Scheme 2). Selective reduction of either carboxylic acid or ester in
compound 8 was attempted. Reduction of ester with LiBH4 was only
successful when a THF solution of 8 and ethanol were added to LiBH4
giving 90% yield. Reduction of carboxylic acid with BH3/THF failed,
while BH3/Me2S gave 70% yield. We abandoned this route due to its
inefficiency and necessity of using expensive palladium catalyst for
subsequent boronylation. Although multi-step masking of acid or
ester groups would allow the use of more economical butyl lithium/
borate method, the lengthened route would be unfavourable for
scale-up. We continued to explore the possibilityof resolving the two
methyl groups in their formyl oxidation state following Route A
(Scheme 1). Slow addition of ethylene glycol to the diformyl com-
pound 3 with azeotropic removal of water gave mono-protected
compound 2 in 62.5% yield (Scheme 2). Subsequent reduction of the
formyl group and protection of the hydroxyl group allowed the use of
butyl lithium/borate method to install boronic acid and the acidic
removal of acetal protection was done in one pot to give 7-for-
mylbenzoxaborole (1) in 44.7% yield over four steps from the difor-
mylbromobenzene (3) (Scheme 2). The ethylene glycol mono-
protection step gave unsatisfactory yield and required column
chromatography, which burdened the large scale synthesis. Fur-
thermore, this syntheticroute for 7-formylbenzoxaborole (1) wasnot
applicable to 6-, 5-, or 4-formylbezoxaboroles because ethylene
COOH
COOMe
8
7
O
O
ethylene glycol
TsOH
Toluene
O
O
CHO
NaBH4
MeOH
Br
Br
Br
OH
CHO
CHO
quant.
CHO
62.5%
3
2
9
1. n-BuLi
2. B(iPrO)3
3. HCl
O
O
OH
MOMCl
Br
B
DIPEA
CH2Cl2
O
OMOM
93.2%
76.7%
1
10
COOH
COOMe
Br
1. SOCl2
Br
Br
KMnO4
100 °C
2. MeOH
0 °C to r.t.
95.5%
t-BuOH/H2O
COOH
COOMe
70 °C
7
6
11
88.3%
OH
Br
LiBH4 /THF
quant.
PCC/Celite
CH2Cl2
CHO
Br
or NaBH4
Dioxane-H2O
78.1%
OH
CHO
92.4%
5
3
Scheme 2. Synthesis of compound 1 and 3.

8740
L. Ye et al. / Tetrahedron 65 (2009) 8738–8744
established a scalable synthetic route with 78% yield over four steps
A
in CD3OD
f
from 2,6-dimethylbromobenzene (6) (Scheme 2). The reactions
were mild and easy-to-control, and high yields allowed rapid puri-
fication without any need of column chromatography, which made
large scale preparation possible. First, 2,6-dimethylbromobenzene
(6) was converted to the dicarboxyl compound 11 using KMnO4 in
t-BuOH/H2O. The yield was as low as 16% under previously reported
reflux condition. We investigated the effects of temperatures and
found that overnight heating at 70 ꢀC gave improved yield (88%)
compared to 110 ꢀC (16%), 100 ꢀC (60%) or 85 ꢀC (77%). At the same
time, slow addition of KMnO4 in portions was essential for satis-
factory yield. To minimize over-oxidation, 2.5 equiv of KMnO4 was
first added slowly and heated for 2 h, then another 2.5 equiv was
added after cooling and followed by heating at 70 ꢀC overnight. The
two carboxyl groups were converted to methyl esters using SOCl2
and CH3OH/Et3N, reduced to hydroxymethyls with LiBH4, then
oxidized to dialdehyde 3 with PCC. All three steps gave near quan-
titative yields that resulted in an overall 78% yield. We also
investigated alternative reducing agents to replace relatively costly
LiBH4. Addition of LiAlH4 gave de-brominated product even at
temperatures as low as ꢁ78 ꢀC. Use of DIBAL gave 81.6% yield.
Reduction with NaBH4 in dioxane/water at 0 ꢀC to room tempera-
ture gave 78.1% yield while reaction of NaBH4 with addition of LiCl or
ZnCl2 failed to produce the desired product. Thus, NaBH4 (dixoane/
water) may be used as an alternative reducing agent with its
reasonable yield and much lower cost.
a
c
e+g
1.11
7.00
2.05
2.00
1.92
4.80
7
6.90
4.70
4.60
a
B
c
in DMSO-d6
d
e/g
f
e/g
0.95
0.93
2.00
5.00
0.99
0.87
2.00
7
5
7.40
7.30
5.50
5.40
5.30 5.10
4.90
4.80
a=c
C
X
in D2O
f
e=g
3.90
0.93
7.50
1.72
5.00
4.90
4.80
7.40
7.30
HO
OH
OH
OH
O
B
B
B(OH)2
OH
B
O
O
O
1
During our study of the H NMR of 7-(30-hydroxypropyl)benz-
H
H
oxaborole (14, Fig. 2), we observed an interesting intramolecular
coordination phenomenon that led to our further investigation and
the development of the final key mono-oxidation step that yielded
an improved synthesis of 7-formylbenzoxaborole (1) (Scheme 3).
This new route is also readily applicable to the synthesis of 6-, 5-
and 4-formylbenzoxaboroles. Compound 14 was synthesized dur-
ing the derivatization of compound 1 (Scheme 4). Wittig reaction
converted compound 1 to the acrylic acid ethyl ester 12, which
underwent hydrogenation using PtO2 to give propionate 13. Re-
duction of the ester with DIBAL gave the hydroxypropyl compound
14 used in the following study.
15
15'
4
A
d
d
O
CD3
D
OH
O
CD3
O
D
OH
c
c
O
OH
B
O
D
B
O
CD3
g
CD3OD
g
B OH
=
O
OH
b
b
slow exchange
f
f
a
e
e
a
d
HO
c
d
O
c
OH
B
OH
g
f
DMSO-d6
x
g
f
B
O
As shown in the 1H NMR of compound 14 in CD3OD, the three
methylene groups give rise to six peaks that can be grouped into two
a
no exchange
b
e
e
a
b OH
0
0
0
sets of signals (Ha, Hb, Hc and Ha , Hb , Hc ), which indicates the ex-
istence of the hydroxypropyl chain in two distinct states (Fig. 2A).
With an empty p-orbital, the trigonal boron can accept a lone elec-
tron pair from heteroatoms X (X¼O, N or F). The B–X bonds can be
strong covalent bonds such as in the boronic acid-based sugar re-
ceptors with an adjacent amine group13 and the bis-methanol ad-
duct of diazaborine14 where boron has a tetrahedral geometry. It can
d
OH
d
O
D
O
OH
D
O
c
c
O
OH
D
D
D
D2O
B
B
O
g
f
g
f
B
OH
=
D
fast exchange
O
O
OH
b
b
a
e
a
e
Figure 3. NMR of compound 15 in CD3OD, DMSO-d6 and D2O, respectively, and the
proposed structures.
OH
CHO
PCC/Celite
CH2Cl2
glycol protection would give mixtures of regioisomers due to their
lack of structural symmetry.
OH
CHO
It is noteworthy that 2,6-diformylbromobenzene (3) is a widely
utilized building block in the construction of various supramole-
cular structures and organometallic catalysts.8–10 Previously, it was
synthesized from 2,6-dimethylbromobenzene by photochemical
tetra-bromination followed by hydrolysis.11 This method involved
vigorous and hard-to-control photochemical reaction, which was
unfavourable for large scale preparation. Another method used
strong oxidants such as Ac2O/CrO3 to convert 2,6-dimethyl-
bromobenzene (6) to 2,6-bis(diacetoxylmethyl)bromobenzene fol-
lowed by acidic hydrolysis to obtain the dialdehyde 3.8,12 It proved
problematic in our attempt to scale it up due to the highly explosive
nature of the oxidant and the low yields in hydrolysis. This
prompted us to explore a new route to circumvent these practical
issues. We developed a milder oxidation–reduction sequence and
17
16
1. n-BuLi
2. B(iPrO)3
3. HCl
HO
OTHP
Br
OH
Br
DHP
TsOH
DMF
OH
B
O
OTHP
OH
quant.
4
5
18
PCC/Celite
CH2Cl2
CHO
OH
B
O
56.1%
2 steps
1
Scheme 3. Improved synthesis of compound 1.

L. Ye et al. / Tetrahedron 65 (2009) 8738–8744
8741
boron (Fig. 3). Interestingly, in DMSO-d6 the sharply resolved signals
indicated no exchange (Fig. 3B). One of the two hydroxyls co-
ordinated to the boron while the other existed in its free form with
the characteristic OH proton as a triplet at 5.48 ppm (Fig. 3B and D).
Proton Ha was downfield shifted to 4.99 ppm, which was charac-
teristic of the well-formed oxaborole ring structure. So compound
15 existed as a well-defined structure of compound 4 in DMSO. This
observation suggested that although both OH groups are positioned
to coordinate to the boron, the bis-coordinated compound 150 was
not formed due to unfavourable constrains of the tricyclic system. In
DMSO-d6, once the stable oxaborole ring was closed, the other OH
group was unable to break it open by exchanging. However, CD3OD
rendered the opening and re-closing of the oxaborole ring feasible.
It has been known that tetrahedral boron forms weaker B–O bond
than trigonal boron due to the loss of conjugation between the
boron p-orbital and the oxygen.16,17 The exchange induced by
CD3OD was likely triggered by coordination of a CD3OD molecule to
the empty p-orbital of boron. This interaction served to induce
a distortion from the trigonal geometry and weaken the B–O bond
that held the oxaborole ring. In D2O, fast exchange was observed as
indicated by the complete merging and sharpening of the aromatic
signals He and Hg as a doublet (Fig. 3C). The two hydroxymethyl
signals Ha and Hc also completely merged into a singlet at 4.93 ppm.
The chemical shift of the merged Ha and Hc in D2O is in between the
coordinated Ha (4.99 ppm) and the free Hc (4.74 ppm) in DMSO-d6
but closer to the coordinated Ha, which suggests that in the fast
exchange process in D2O, the averaged structure has both hydroxyls
coordinated to boron to a significant extent (Fig. 3).
The above intramolecular coordination properties indicated that
compound 15 existed as a well-defined structure of compound 4 in
DMSO, but had slow to fast exchange in coordinating solvent
CD3OD or D2O. This observation suggested that in non-coordinating
solvents, one of the two hydroxyls could be protected by forming
a strong intramolecular B–O bond while leaving the other hydroxyl
to be manipulated by chemical modifications. We tested the as-
sumption by carrying out the oxidation of compound 4 with excess
PCC in methylene chloride and obtained mono-oxidized compound
1 in quantitative yield (Scheme 3). On the contrary, 1,3-bis(hy-
droxymethyl)benzene 16, which lacks the boronic acid gave
1,3-diformylbenzene 17 as the main product under the same con-
dition, which is consistent with previous reports where dialdehyde
was the major product with even 1 equiv of oxidant18 and where
low yield (68%) was obtained with insufficient amount of PCC.19
This efficient mono-oxidation allowed us to shorten the synthesis
of compound 1 (Scheme 3). Compound 5 was converted to THP
protected compound 18 in quantitative yield. Boronic acid was in-
stalled using n-BuLi and B(i-PrO)3, and HCl was added subsequently
to remove THP groups, producing compound 4. Due to its high
water solubility thus postulated problem in purification, crude
compound 4 was directly oxidized with PCC to compound 1. The
final product was purified by taking advantage of the Lewis acidity
of the boronic ester group. At basic pH, with the coordination of
two hydroxide ions the boron was negatively charged thus
remained in aqueous phase, which allowed the removal of impu-
rities into organic phase. Subsequently, it was acidified to neutral
form and extracted into ethyl acetate to give 56.1% yield over two
steps with satisfactory purity. This improved route gave the desired
7-formylbenzoxaborole (1) in 47.3% overall yield over six steps from
commercially available 2,6-dimethylbromobenzene (6). More im-
portantly, it was proved to be amenable to large scale synthesis
with its high yields, easy-to-control reaction conditions, robust
reproducibility and facile purification without any need of column
chromatography.
O
O
CHO
OH
OH
B
B
O
O
19
NaOH, EtOH/H2O
1
EtO
O
CHO
OH
OH
B
B
Ph3PCH2COOCH2CH3Br
NaH/THF
O
O
1
12
EtO
O
HO
PtO2-3H2O
H2
DIBAL
OH
OH
B
B
O
O
14
13
Bz
O
N
CHO
OH
OH
B
B
PhCH2ONH2
O
O
20
1
HO
N
OH
CHO
OH
HO
B
B
NH
O
O
NaBH(OAc)3
1
21
Scheme 4. Representative reactions of compound 1.
also be an intermediate B–X interaction that is weaker than covalent
bond but has detectable electron effects resulting in a more shielded
boron such as in the bis(boronate-amide) receptor.15 Thus, we de-
duced two coordination states as shown in Figure 2C to account for
the two sets of methylene signals. Protons Ha, Hb and Hc are from the
free hydroxypropyl chain while protons Ha , Hb and Hc , with
downfield shifts, result from the coordination of the terminal hy-
droxyl to boron. The flow of electron from hydroxyl oxygen to boron
0
0
0
0
0
0
resulted in the more de-shielded Ha , Hb and Hc . Furthermore, we
added metallic sodium to replace the terminal hydroxyl group with
high affinity CD3Oꢁ anion. Indeed, only a single component was
observed that corresponded to a free hydroxypropyl chain (Fig. 2B).
This is the first time an intramolecular reversible coordination in
benzoxaborole is reported. The five-membered oxaborole ring is
stable and spontaneously formed.4–7 Although there were two ter-
minal hydroxyls in compound 14, only the hydroxypropyl–OH was in
reversible coordination with boron while the hydroxymethyl–OH
was locked by the stable oxaborole ring. Benzoxaborole, as a cyclic
boronic ester, has properties rather different from acyclic aromatic
boronic ester. In compound A (Fig. 3), the five-membered oxaborole
ring is rather stable and stays closed in DMSO, methanol and water as
indicated by the constant chemical shifts (w5.0 ppm) of the meth-
ylene protons. We were intrigued to ask what would happen if there
were two equivalent hydroxymethyl groups adjacent to boronic acid
as illustrated by compound 15 (Fig. 3).
Compound 15 was obtained by reduction of compound 1 with
NaBH4 for the subsequent study. The 1H NMR of compound 15 in
CD3OD showed a slow exchange as indicated by the broadened
signal of two aromatic protons at 6.91–6.97 ppm (Fig. 3A). We
assigned it to He and Hg and the merging of the two protons can be
accounted for by the slow exchange of Ob or Od coordinating to
It is noteworthy that this new route is readily applicable to the
synthesis of 6-, 5- or 4-formylbenzoxaboroles as further demon-
strated in the synthesis of 6-formylbenzoxaborole (22) from

8742
L. Ye et al. / Tetrahedron 65 (2009) 8738–8744
2,5-dimethylbromobenzene (24) (Scheme 5). This, for the first time,
provided an opportunity to rapidly diversify the under-explored
benzoxaboroles as medicinally important molecules. We demon-
strated the derivatization of these formylbenzoxaboroles under
conditions compatible with the stability of the oxaborole ring by
(400 MHz, DMSO-d6):
d
13.58 (s, 2H), 7.68 (m, 2H), 7.50 (q, J1¼8 Hz,
J2¼7.2 Hz, 4H); 13C NMR (100 MHz, DMSO-d6):
d 168.1, 137.1, 131.1,
128.2, 116.6; HRMS-EI: C8H5BrO4 calcd 243.9371, found 243.9372;
mp: 218–221 ꢀC.
the representative synthesis of
a,b-unsaturated ketone (19) by aldol
4.2. Preparation of 2-bromo-isophthalic acid dimethyl
ester (7)
condensation, -unsaturated ester (12) by Wittig reaction, O-
a,b
alkyloxime (20) by condensation with O-alkyl hydroxylamine,
amine (21) by reductive amination (Scheme 4). The benzoxaborole
moiety, known to be unstable under certain oxidative or alkaline
conditions due to C–B bond cleavage, proved to be intact in these
reactions.
Compound 11 (44.20 g, 180.4 mmol) in SOCl2 (300 mL) was
gradually heated to 100 ꢀC for a period of 5 h and stirred at 100 ꢀC
for another 4 h. After SOCl2 was evaporated in vacuo and the flask
was cooled to 0 ꢀC, methanol (200 mL) and triethylamine (100 mL)
were added slowly while stirring. The reaction mixture was stirred
at room temperature for 2 h and was concentrated in vacuo. The
residue was extracted with EtOAc, dried over MgSO4 and concen-
trated in vacuo to obtain 47.05 g of compound 7 (95.5%) as viscous
OH
OH
OH
Br
OHC
B
B
O
O
oil. 1H NMR (400 MHz, CDCl3):
6H); 13C NMR (100 MHz, CDCl3):
d
7.70 (m, 2H), 7.40 (t, 1H), 3.94 (s,
166.0, 134.8, 132.0, 126.9, 117.9,
22
23
24
d
52.5; HRMS-EI: C10H9BrO4 calcd 241.9684, found 241.9683.
Scheme 5. Synthesis of 6-formyl benzoxaborole.
4.3. Preparation of 2,6-bis(hydroxymethyl) bromobenzene (5)
3. Conclusions
To a solution of compound 7 (30.50 g, 112 mmol) in ethyl ether
(300 mL) was added LiBH4 (5.42 g, 245.8 mmol) in THF (100 mL)
slowly at 0 ꢀC. The reaction mixture was stirred at room tempera-
ture overnight, quenched with 0.4 M HCl to pH¼6w7 and extracted
with ethyl acetate to obtain 24.51 g of compound 5 (quantitative).
Alternative method: compound 7 (13.65 g, 50.0 mmol) was
dissolved in 250 mL 1,4-dioxane/H2O (3:2, 250 mL) and cooled to
0 ꢀC. To this mixture was added NaBH4 (18.90 g, 0.50 mol) and
stirred at room temperature for 2 d before it was quenched with
6 M HCl in ice-bath, extracted with ethyl acetate, washed with
saturated NaHCO3 and brine, dried over Na2SO4 and concentrated
in vacuo to obtain 8.50 g of compound 5 (78.1%).
Benzoxaborole AN2690 (Fig. 1) was recently identified as
a novel aminoacyl tRNA synthetase inhibitor3 and is currently
under clinical trials as a treatment of onychomycosis.4 However,
synthetic method for rapid structural diversification of benzox-
aborole scaffold is largely under-explored despite of their me-
dicinal significance. We developed a practically scalable synthesis
of the versatile 7-formylbenzoxaborole (1) that will allow rapid
derivatization and exploration of their medicinal value in depth.
We obtained a 47.3% overall yield over six steps from the com-
mercially available 2,6-dimethylbromobenzene (6). Furthermore,
this method is readily applicable to the synthesis of all three other
regioisomers, 4-, 5- and 6-formylbenzoxaboroles. The application
to the synthesis of 6-formylbenzoxaborole was also carried out
and the derivatization reactions were demonstrated to be com-
patible with the benzoxaborole stability. The key mono-oxidation
step was developed based on the observation that the strength of
the intramolecular B–O interaction relies on ring size and co-
ordinating ability of the solvent. We also developed a mild and
scalable synthesis of the intermediate 1,3-diformylbromobenzene
(3), which has wide use in supramolecular and organometallic
chemistry. Owing to the recently highlighted significance of
benzoxaboroles in drug discovery and the rarity of a systematic
synthesis and modification strategy, we believe this method will
find wide application in the field of drug discovery and biological
mechanistic studies.
1H NMR (400 MHz, DMSO-d6):
d
7.39 (m, 3H), 5.39 (t, J¼5.6 Hz,
2H), 4.52 (d, J¼5.6 Hz, 4H); 13C NMR (100 MHz, DMSO-d6):
d 141.0,
127.0, 126.3, 120.4, 62.9; HRMS-EI: C8H9BrO2 calcd 215.9786, found
215.9783; mp: 167–169 ꢀC.
4.4. Preparation of 2-bromobenzene-1,3-dicarbaldehyde (3)
A mixture of compound 5 (12.0 g, 55.3 mmol), PCC (35.7 g,
165.9 mmol) and Celite (53.6 g) in CH2Cl2 (500 mL) was stirred at
room temperature overnight. The reaction mixture was filtered
through Celite and silica gel pad and the filtrate was evaporated in
vacuo to obtain 10.90 g of compound 3 (92.4%). 1H NMR (400 MHz,
DMSO-d6):
(75 MHz, CDCl3):
d
10.37 (s, 2H), 8.08 (m, 2H) and 7.71 (m, 1H); 13C NMR
d
190.7, 135.3, 134.4, 130.9, 128.3; HRMS-EI
C8H5BrO2 calcd 211.9473, found 211.9472; mp: 137–141 ꢀC.
4. Experimental
4.5. Preparation of 2-(20-bromo-30-((tetrahydro-pyran-200-
yloxy)methyl)benzyloxy)-tetrahydro-pyran (18)
4.1. Preparation of 2-bromo-1,3-benzenedicarboxylic
acid (11)
Compound 5 (18.0 g, 82.9 mmol) and p-toluenesulfonic acid
monohydrate (0.78 g, 4.1 mmol) were dissolved in DMF (200 mL),
and DHP (34.8 mL, 331.7 mmol) was added. The reaction mixture
was stirred at room temperature for 1 d and was quenched with
saturated NaHCO3, extracted with ethyl ether, washed with brine,
dried over Na2SO4 and evaporated in vacuo to obtain 34.15 g of
compound 18 (quantitative) as a viscous oil. 1H NMR (400 MHz,
To a solution of compound 6 (60 g, 324.3 mmol) in t-BuOH
(250 mL) and H2O (250 mL) was added KMnO4 (128 g, 0.81 mol) in
portions while stirring at room temperature. The mixture was
stirred at 70 ꢀC for 2 h before it was cooled to room temperature. A
second batch of KMnO4 (128 g, 0.81 mol) was added as before. After
stirring at 70 C for 10 h, the hot reaction mixture was filtered and
the residue was washed with water (3ꢂ300 mL). After concentra-
tion to 300 mL, the filtrate was acidified in ice-bath to pH¼2 with
concd HCl to get white precipitate. After extraction with EtOAc
(4ꢂ500 mL), the organic phase was dried with Na2SO4 and con-
centrated in vacuo to obtain 70.2 g of compound 11 (88.3%). 1H NMR
CDCl3):
J1¼33.2 Hz, J2¼13.2 Hz, 2H), 3.89 (m, 2H), 3.53 (m, 2H), 1.94 (m,
12H); 13C NMR (100 MHz, CDCl3):
137.9, 127.6, 126.9, 122.6, 98.1,
d 7.44 (m, 2H), 7.32 (m, 1H), 4.86 (m, 2H), 4.72 (dd,
d
68.5, 61.8, 30.3, 25.2, 19.1; HRMS-EI: C18H25BrO4 calcd 384.0936,
found 384.0935.

L. Ye et al. / Tetrahedron 65 (2009) 8738–8744
8743
4.6. Preparation of 7-formyl-1-hydroxy-1,3-dihydro-2,1-
benzoxaborole (1)
71.9,61.5,14.6;HRMS-EI:C12H13BO4calcd232.0907,found232.0909;
mp:151–152 ꢀC.
Compound 18 (14.0 g, 36.3 mmol) was dissolved in THF
(200 mL) and n-BuLi (1.6 M in hexane, 25 mL, 40 mmol) was added
dropwise at ꢁ78 ꢀC. The reaction mixture was stirred at ꢁ78 ꢀC for
20 min and triisopropyl borate (9.2 mL, 40 mmol) was added. The
reaction mixture was allowed to warm to room temperature and
stirred overnight before it was quenched with 6 M HCl (100 mL),
stirred at room temperature for 6 h and concentrated in vacuo. The
crude compound 4 was directly dissolved in CH2Cl2 (150 mL), and
PCC (15.7 g, 72.6 mmol) and Celite (23.5 g) were added. The mix-
ture was stirred at room temperature overnight and was filtered
through a Celite and silica pad and the filtrate was washed with
1 M HCl and extracted with 1 M NaOH. The aqueous phase was
acidified with concd HCl to pH¼2 and extracted with ethyl acetate.
The resulting organic phase was washed with brine, dried with
Na2SO4 and evaporated in vacuo to obtain 3.30 g of compound 1
4.9. Preparation of [3-(1,3-dihydro-1-hydroxy-2,1-
benzoxaborol-7-yl)]propionic acid ethyl ester (13)
To a solution of compound 12 (1.20 g, 5.17 mmol) in methanol
(10 mL) was added PtO2$3H2O (73 mg, 0.26 mmol) and the reaction
mixture was vacuumed and filled with hydrogen and stirred
overnight at room temperature. The reaction mixture was filtered
and evaporated in vacuo to obtain 1.16 g of compound 13 (96%). 1H
NMR (400 MHz, CD3OD):
d
7.34 (t, J¼7.6 Hz, 1H), 7.18 (d, J¼7.6 Hz,
1H), 7.11 (d, J¼7.2 Hz, 1H), 5.02 (s, 2H), 4.06 (q, J¼7.2 Hz, 2H), 3.05 (t,
J¼7.6 Hz, 2H), 2.61 (t, J¼7.6 Hz, 2H) and 1.19 (t, J¼7.2 Hz, 3H); 13C
NMR (100 MHz, CD3OD):
d 174.9, 132.2, 128.2, 120.1, 72.1, 61.4, 37.1,
30.9, 14.5; HRMS-EI: C12H15BO4 calcd 234.1063, found 234.1060;
mp: 93–95 ꢀC.
(56.1% over two steps). 1H NMR (400 MHz, DMSO-d6):
1H), 9.22 (s, 1H), 7.87 (m, 1H), 7.72 (m, 2H), 5.13 (s, 2H); 13C NMR
(100 MHz, CD3OD): 155.1, 142.9, 132.2, 131.9, 125.2, 122.4, 104.0,
72.0; HRMS-EI C8H7BO3 calcd 162.0488, found 162.0490; mp: 146–
148 ꢀC. Compound 4: 1H NMR (400 MHz, DMSO-d6):
8.9 (s, 1H),
d
10.39 (s,
4.10. Preparation of 7-(30-hydroxypropyl)-1,3-dihydro-1-
hydroxy-2,1-benzoxaborole (14)
d
To a solution of compound 13 (500 mg, 2.15 mmol) in THF
(20 mL) was added dropwise DIBAL (1.0 M in hexane, 12.9 mL,
12.9 mmol) at 0 ꢀC. The reaction mixture was stirred overnight at
room temperature and quenched with 1 M HCl at 0 ꢀC. The mixture
was extracted with ethyl acetate, washed with brine and dried over
Na2SO4. After rotary evaporation, the residue was purified by col-
umn chromatography to obtain 270 mg of compound 14 (65.5%). 1H
d
7.44 (t, J¼7.2 Hz, 1H), 7.33 (d, J¼7.2 Hz, 1H), 7.25 (d, J¼7.2 Hz, 1H),
5.48 (t, J¼5.6 Hz, 1H), 4.99 (s, 2H) and 4.74 (d, J¼5.6 Hz, 2H); 13C
NMR (75 MHz, D2O):
d
131.8, 123.1 (br), 66.7 (br); 13C NMR
(100 MHz, D2OþNaOH):
d 148.1, 141.2, 127.2, 124.6, 120.2, 67.5,
64.1; HRMS-EI: C8H9BO3 calcd 164.0645, found 164.0650; mp:
290–291 ꢀC.
NMR (400 MHz, CD3OD, Na added):
6.83 (m, 1H), 4.81 (s, 2H), 3.48 (m, 2H), 2.77 (m, 2H) and 1.84 (m,
2H); 13C NMR (100 MHz, CD3OD):
155.5, 147.9, 132.5, 132.1, 128.1,
d 7.03 (m, 1H), 6.91 (m, 1H),
d
4.7. Preparation of (E)-[3-(1,3-dihydro-1-hydroxy-2,1-
benzoxaborol-7-yl)-1-phenyl]propenone (19)
127.1, 119.6, 72.8, 72.1, 70.5, 62.5, 37.0, 35.9, 32.0, 31.8; HRMS-EI
C10H13BO3 calcd [Mꢁ1]ꢁ 191.0879, found 191.0882; mp: 91–93 ꢀC.
To a mixture of acetophenone (0.22 mL, 1.85 mmol), ethanol
(5 mL) and water (8 mL) was added NaOH (296 mg, 7.41 mmol).
After stirring for 5 min compound 1 (300 mg, 1.85 mmol) was
added and the reaction mixture was stirred at room temperature
overnight before quenched with 6 M HCl to pH¼2 in ice-bath. The
mixture was evaporated, extracted with ethyl acetate and dried
over anhydrous Na2SO4. The residue after rotary evaporation was
purified by column chromatography and re-crystallization (hexane/
ethyl acetate) to obtain 240 mg of compound 19 (49.1%). 1H NMR
4.11. Preparation of (E)-(1,3-dihydro-1-hydroxy-2,1-
benzoxaborol-7-yl)-O-benzyl oxime (20)
To a mixture of compound 1 (200 mg, 1.24 mmol), O-benzyl-
hydroxylamine hydrochloride (240 mg, 1.50 mmol) and sodium
formate (0.48 g, 7.1 mmol) was added 88% formic acid (1.9 mL). The
mixture was heated at 85 ꢀC overnight, quenched with water
(10 mL) and extracted with ethyl acetate (20 mLꢂ3). The residue
after rotary evaporation was purified by column chromatography to
obtain 222 mg of compound 20 (59.3%). 1H NMR (300 MHz, DMSO-
(300 MHz, DMSO-d6):
d
9.39 (s, 1H), 8.14 (m, 5H), 7.68 (t, J¼7.4 Hz,
1H), 7.58 (t, J¼8 Hz, 3H), 7.48 (m, 1H) and 5.05 (s, 2H); 13C NMR
d6):
d
9.06 (s, 1H), 8.58 (s, 1H), 7.67 (d, J¼5.4 Hz, 1H), 7.51 (t,
(300 MHz, CD3OD)
d 192.8, 156.2, 145.3, 139.6, 139.4, 134.2, 132.4,
J¼5.4 Hz, 1H), 7.39 (m, 6H), 5.19 (s, 2H) and 5.02 (s, 2H); 13C NMR
129.7, 127.4, 124.8, 123.9, 71.9; HRMS-EI C16H13BO3 calcd 264.0958,
found 264.0964; mp: 136–137 ꢀC.
(100 MHz, CDCl3):
d 155.2, 152.6, 136.3, 133.7, 131.3, 130.1, 128.8,
128.7, 128.6, 123.3, 76.9, 71.5; HRMS-EI C15H14BNO3 calcd 267.1067,
found 267.1069; mp 56–58 ꢀC.
4.8. Preparation of (E)-[3-(1,3-dihydro-1-hydroxy-2,1-
benzoxaborol-7-yl)]acrylic acid ethyl ester (12)
4.12. Preparation of 7-{[4-(hydroxymethyl)piperidin-1-
yl]methyl}-1-hydroxy-1,3-dihydro-2,1-benzoxaborole (21)
To a solution of compound 1 (400 mg, 2.47 mmol) and (ethyl-
oxycarbonylmethyl)triphenylphosphonium
bromide
(1.06 g,
To a solution of compound 1 (320 mg, 1.98 mmol) and (piper-
idin-4-yl) methanol (227 mg, 1.98 mmol) in 1,2-dichoroethane
(10 mL) in ice-bath was added NaBH(OAc)3 (587.5 mg, 2.77 mmol).
The reaction mixture was stirred at room temperature overnight,
quenched with saturated NaHCO3 and washed with ethyl acetate.
The aqueous phase was concentrated and purified by reversed
phase column chromatography to obtain 173 mg of compound 21
2.47 mmol)inTHF(25 mL)wasadded under stirring NaH (60% in oil,
99 mg,2.47 mmol)atꢁ5 ꢀC.Thereactionmixturewasstirredatroom
temperature for 12 h, added another portion of NaH (50 mg,
1.24 mmol)at0 ꢀCandstirredatroomtemperaturefor8 h.Thereaction
wasquenchedwithwater,acidifiedtopH¼2w3,extractedwithethyl
acetateanddriedoverNa2SO4.Theresidueafterrotaryevaporationwas
purified bycolumn chromatographyand re-crystallization (hexane/
ethylacetate)toobtain200 mgofcompound12(34.9%yield).1HNMR
(173 mg, 33.7%). 1H NMR (400 MHz, CD3OD):
d 7.27 (m, 2H), 7.17 (m,
1H), 4.83 (s, 2H), 4.29 (s, 2H), 3.48 (m, 4H), 2.93 (m, 2H), 1.96 (m,
(300 MHz, DMSO-d6):
d
9.33 (s, 1H), 8.10 (d, J¼16.2 Hz, 1H), 7.82 (d,
1H), 1.82 (m, 2H) and 1.38 (m, 2H); 13C NMR (100 MHz, CD3OD):
J¼10 Hz,1H),7.52(t,J¼7.5 Hz,1H),7.44(m,1H),6.81(d,J¼16.2 Hz,1H),
d 150.8, 132.4, 128.3, 128.0, 122.7, 71.1, 66.5, 62.7, 52.8, 37.6, 27.7;
5.02 (s, 2H), 4.19 (q, J¼7.1 Hz, 2H) and 1.26 (t, J¼7.2 Hz, 3H); 13C NMR
HRMS-EI C14H20BNO3 calcd [M-1]- 260.1458, found 260.1461; mp:
109–113 ꢀC.
(300 MHz,CD3OD):d168.8,155.9,144.9,138.9,132.3,125.9,123.7,120.1,

8744
L. Ye et al. / Tetrahedron 65 (2009) 8738–8744
3. Rock, F. L.; Mao, W.; Yaremchuk, A.; Tukalo, M.; Crepin, T.; Zhou, H.; Zhang,
Y.-K.; Hernandez, V.; Akama, T.; Baker, S. J.; Plattner, J. J.; Shapiro, L.; Mar-
tinis, S. A.; Benkovic, S. J.; Cusack, S.; Alley, M. R. K. Science 2007, 316, 1759–
1761.
4. Baker, S. J.; Zhang, Y.-K.; Akama, T.; Lau, A.; Zhou, H.; Hernandez, V.; Mao, W.;
Alley, M. R. K.; Sanders, V.; Plattner, J. J. J. Med. Chem. 2006, 49, 4447–4450.
5. Tschampel, P.; Snyder, H. R. J. Org. Chem. 1964, 29, 2168–2172.
6. Gunasekera, D. S.; Gerold, D. J.; Aalderks, N. S.; Chandra, J. S.; Maanu, C. A.;
Kiprof, P.; Zhdankin, V. V.; Reddy, M. V. R. Tetrahedron 2007, 63, 9401–9405.
7. Akama, T.; Baker, S. J.; Zhang, Y. K.; Hernandez, V.; Zhou, H.; Sanders, V.; Freund,
Y.; Kimura, R.; Maples, K. R.; Plattner, J. J. Bioorg. Med. Chem. Lett. 2009, 19,
2129–2132.
4.13. Preparation of 6-formyl-1-hydroxy-1,3-dihydro-2,1-
benzoxaborole (22)
Procedure similar to that of compound 1. 1H NMR (300 MHz,
DMSO-d6):
d
10.06 (s, 1H), 9.44 (s, 1H), 8.28 (s, 1H), 8.01 (m, 1H),
194.1,
7.63 (m, 1H) and 5.09 (s, 2H); 13C NMR (100 MHz, CD3OD):
d
133.6, 132.5, 130.6, 129.6, 123.2, 122.1, 72.2; HRMS-EI: C8H7BO3
calcd 162.0488, found 162.0494; mp: 133–135 ꢀC.
8. Steffen, K.; Thomas, G.; Berthold, K. Angew. Chem., Int. Ed. 2005, 45, 101–104.
9. Fossey, J. S.; Russell, M. L.; Abdul Malik, K. M.; Richards, C. J. J. Organomet. Chem.
2007, 692, 4843–4848.
Acknowledgements
10. Oakley, S. H.; Coogan, M. P.; Arthur, R. J. Organometallics 2007, 26, 2285–2290.
11. Harvey, R. G.; Dai, Q.; Ran, C.; Penning, T. M. J. Org. Chem. 2004, 69, 2024–2032.
12. Wille, E. E.; Stephenson, D. S.; Capriel, P.; Binsch, G. J. Am. Chem. Soc. 1982, 104,
405–415.
13. Wiskur, S. L.; Lavigne, J. J.; Ait-Haddou, H.; Lynch, V.; Chiu, Y. H.; Canary, J. W.;
Anslyn, E. V. Org. Lett. 2001, 3, 1311–1314.
14. Groziak, M. P.;Ganguly, A. D.; Robinson, P. D. J. Am. Chem. Soc.1994,116, 7597–7605.
15. Hughes, M. P.; Smith, B. D. J. Org. Chem. 1997, 62, 4492–4499.
16. Matteson, D. S.; Michnick, T. J.; Willett, R. D.; Patterson, C. D. Organometallics
1989, 8, 726–729.
17. Matteson, D. S. In Stereodirected Synthesis with Organoboranes; Rees, C. W., Lehn,
J. M., Zahradnik, R., Hafner, K., Eds.; Springer: Berlin, 1995; pp 1–20.
18. Constantinides, I.; Macomber, R. S. J. Org. Chem. 1992, 57, 6063–6067.
19. Boge, T. C.; Wu, Z. J.; Himes, R. H.; Velde, D. G. V.; Georg, G. I. Bioorg. Med. Chem.
Lett. 1999, 9, 3047–3052.
We thank National Natural Science Foundation of China
(20702031), Shanghai Municipal Education Commission (07SG15),
Science and Technology Commission of Shanghai Municipality
(07pj14095), Ministry of Science and Technology of China
(2009CB918404) and Anacor Pharmaceuticals for financial support
of this work.
References and notes
1. Yang, W.; Gao, X.; Wang, B. In Boronic Acids; Hall, D. G., Ed.; Wiley-VCH:
Weinheim, 2005; pp 481–512.
2. Paramore, A.; Frantz, S. Nat. Rev. Drug Discov. 2003, 2, 611–612.